What You See is What You Get: Distributional Generalization for Algorithm Design in Deep Learning. (arXiv:2204.03230v1 [cs.LG])

We investigate and leverage a connection between Differential Privacy (DP)
and the recently proposed notion of Distributional Generalization (DG).
Applying this connection, we introduce new conceptual tools for designing
deep-learning methods that bypass "pathologies" of standard stochastic gradient
descent (SGD). First, we prove that differentially private methods satisfy a
"What You See Is What You Get (WYSIWYG)" generalization guarantee: whatever a
model does on its train data is almost exactly what it will do at test time.
This guarantee is …

algorithm algorithm design arxiv deep learning design learning